The goal of this proposal is to develop antibody library-based methods which promote efficient identification of clinically relevant tumor specific cell surface antigens, methods which are applicable to the identification of cell type-specific lineage markers in general. The abnormal physiology of tumor cells is reflected in part in the altered chemical and molecular composition of their cell surface. Progressive changes in surface molecule expression allow tumor cells to respond efficiently to external signals for growth and survival, to interact with host tissues, to achieve metastasis, and to avoid immune surveillance. The identification and targeting of tumor-specific cell surface antigens, however, is hampered by the complexity of the epitope space at the tumor cell surface. In addition to proteins, relevant antigens include carbohydrates and other post-translational modification products that cannot be predicted from studies of genomic copy number or mRNA expression levels. This proposal aims to develop combinatiroal antibody library-based strategies that allow efficient identification of tumor specific cell surface antigens for diagnostic, prognostic and therapeutic applications. Specifically, we aim (1) To develop a high throughput subtractive selection strategy based on flow cytometric sorting to significantly improve selection efficiency and to obtain greater numbers of tumor- targeting phage antibodies. (2) To identify clinically relevant tumor antigens by selecting phage antibodies against prostate cancer cells in situ, within their proper stromal contexts. We propose to combine antibody library technology with laser capture microdissection (LCM) to identify antibodies against individual tumor cells on tissue slides. The resulting antibodies will have a very high likelihood of recognizing clinically relevant tumor antigens. (3) To establish the molecular identify of tumor cell surface antigens recognized by tumor-specific antibodies. We have developed a novel combinatorial yeast cDNA display library. Specific antigen-antibody pairs will be identified from this library following flow sorting. In parallel, we will pursue the analysis of tumor antigens by mass spectrometry methods which are capable of analyzing post-translational modifications. This study will increase our knowledge of tumor physiology and will facilitate the design of effective therapy; it will also help determine the exact contribution of post-translational modifications to the final makeup of the tumor epitope space.